Posts from machine learning

Machine Learning For Beginners

So you are hearing a lot about machine learning these days.

You are hearing words like models, training, forks, splits, branches, leafs, recursion, test data, and overfitting, and you don’t know what any of them mean.

Well I have some help, courtesy of my colleague Jacqueline who shared this scrolling lesson in machine learning with her colleagues at USV (me included) this weekend.

This scrolling machine learning lesson was made by Stephanie and Tony. It is great work. Thanks!

Brain Computer Interface

The WSJ reported yesterday that Elon Musk is developing yet another company, this one based on neural lace technology, to create a brain computer interface.

Neural lace technology, as I understand it, involves implanting electrodes into the brain so that the brain can control machines directly without the need for an IO device like a mouse, keyboard, or voice interface.

I have no idea how advanced this technology is and whether it is ready for commercialization or if this is basically a research project masquerading as a startup.

But in some ways that doesn’t matter if you believe that at some point someone or some group of scientists and medical professionals will figure out how to directly connect our brain to machines without the need of an IO device.

There are so many times that I have thoughts that I don’t do anything with. They sit idle and maybe go nowhere. But if my brain passively passed those thoughts onto a machine for storage or some other action that could lead to a more productive train of thought that could be incredibly valuable. Or it could drive me insane.

I generally subscribe to the theory that all progress is good as long as we understand the negatives of the technology and we (society) engineer controls and the proper repoanes to it (nuclear weapons being​ an example).

But every time something as mind bending as the idea of connecting our brain to external processing, storage, and communication infrastructure comes before me I do have to pause and ask where this is all going.

At times like this it helps to have a belief system (progress is good). I am all for pushing the envelope of progress as long as we spend an equal amount of time and energy thinking through what might go wrong with things like this.

Hat tip to Niv Dror who read yesterday that I wasn’t sure how I was going to post today and encouraged me to write about this topic.

AI For Legal Cases

Our portfolio Casetext was in the news yesterday for raising $12mm, but the more interesting thing about Casetext is their product, called CARA.

CARA is a research assistant for lawyers that offers a super simple proposition:

Securely upload a brief and discover useful case law

CARA uses Casetext’s wikipedia-like database of >10mm court cases and annotations and sophisticated natural language analysis and artificial intelligence to understand the brief and recommend related cases for a lawyer to analyze and possibly cite in their brief.

Lawyers seem to love CARA. According to Silicon Valley Business Journal:

Casetext’s customers include Quinn Emanuel, Fenwick & West, Ogletree Deakins, Greenberg Traurig and DLA Piper.

“CARA is an invaluable, innovative research tool,” Quinn Emanuel partner David Eiseman said in a statement. “With CARA, we can upload a brief and within seconds receive additional case law suggestions and relevant information on how cases have been used in the past, all in a user-friendly interface.”

We think the legal business is ripe for AI-driven innovation. Much of legal research can and will be automated with tools like CARA.

If you are a lawyer and do a lot of legal research, check out CARA. Securely upload a brief here and check it out.

Machine Learning For Investing In Consumer Goods Startups

Our portfolio company CircleUp has been building a marketplace for startup investing, by accredited and institutional investors, in consumer goods companies (natural foods, personal care, beverage, home goods and apparel). In four years of operation, over $300mm has been raised on CircleUp by entrepreneurs to scale their consumer goods startups.

But underneath all of this has been a sophisticated data science effort designed to track the entire consumer goods sector (all companies, not just the ones on CircleUp) and determine which companies succeed and why. Yesterday CircleUp took the covers off this data science effort, called Helio, and explained what they are up to with it.

Here are some bits from that blog post:

there’s endless data on consumer product and retail companies. And, much of it is public. A quick Google search of the product in your pantry tells you how many SKUs the brand has, price points for each SKU, where they are sold, product reviews, and a great deal more. In an A16Z podcast in 2016, Marc Andreessen commented that machine learning wouldn’t be helpful for tech VC because there isn’t enough data (40:04 mark). We agree. But in the consumer industry, the opposite is true. Data is broadly available. Business models are uniform. That’s the perfect recipe for machine learning. That makes Helio possible.

Let’s take a look at a few examples:

  • Supergoop! is a sunscreen brand available nationally throughout Sephora, that Helio surfaced due to its quickly growing brand, great distribution and estimated revenue growth. We presented Supergoop! to institutional investors, and shortly after, they raised $3.25 million.
  • REBBL is a line of coconut-milk based beverages made with super herbs known to reduce stress. Aside from being one of the fastest growing products in its category, REBBL donates 2.5% of net sales to initiatives helping eradicate human trafficking. Helio spotted REBBL early and qualified it for investors, showing its compelling brand, team and distribution metrics. Today, REBBL’s lead investors include Powerplant Ventures, led by the ZICO coconut water founder, and Boulder Investment Group Reprise.
  • nutpods plays in the crowded plant-based, dairy alternative category. Helio spotted nutpods for its remarkable product reviews, strong early growth and overall brand, despite it having less than $50,000 in annual sales at the time. After, nutpods got investments from Stray Dog Capital and Melissa Hartwig, founder and CEO of Whole 30, and today is rated #1 on Amazon in its category.
  • Tio Gazpacho is a quickly growing brand in the relatively new category of bottled soups, or more broadly, drinkable meals. Tio Gazpacho was founded in Florida, a place without a robust VC community, but Helio still spotted it, and surfaced it to General Mills, which now is its lead investor.

Helio is currently monitoring over a million brands across natural foods, personal care, beverage, home goods and apparel, and can help find who might be the next Krave Jerky, Seventh Generation or Too Faced. We are talking to likely candidates right now, and not just in the categories above, in all categories we see as promising growth areas in the consumer market.

CircleUp has always taken the view that the entrepreneurs with the best ideas, products and team should win…not the one with the best personal connections. Helio brings us a big step closer towards that ambition.

We are excited to see what happens when entrepreneurs with big ideas meet a capital market that has data science at the core. If you want to participate in that market, visit CircleUp.

The Robot Tax And Basic Income

In my work to prepare for the Future of Labor conversation we had at NewCo Shift a few weeks ago, I talked to a number of experts who are studying job losses due to automation and thinking about what might be done about it. Two ideas that came up a number of times were the “robot tax” and the “basic income.”

The ideas are complementary and one might fund the other.

At its simplest, a “robot tax” is a tax on companies that choose to use automation to replace human jobs. There are obviously many variants of this idea and to my knowledge, no country or other taxing authority has implemented a robot tax yet.

A “basic income” is the idea that everyone receives enough money from the government to pay for their basic needs; housing, food, clothing so that as automation puts people out of work we don’t see millions of people being put out on the street.

What is interesting about these two ideas is that some of the biggest proponents of them are technology entrepreneurs and investors, the very people who are building and funding the automation technologies that have the potential to displace many jobs.

It is certainly true that we don’t know that automation will lead to a jobs crisis. Other technological revolutions like farming and factories produced as many new jobs as they wiped out and incomes increased from these changes. Automation could well do the same.

But smart people are wondering, both privately and publicly, if this time may be different. And so ideas like the robot tax and the basic income are getting traction and are being studied and promoted.

The latest proponent of a robot tax is Bill Gates who said this about it:

You ought to be willing to raise the tax level and even slow down the speed. That’s because the technology and business cases for replacing humans in a wide range of jobs are arriving simultaneously, and it’s important to be able to manage that displacement. You cross the threshold of job replacement of certain activities all sort of at once.

There is a lot of economic surplus that could come from automation. Let’s look at ride sharing. Today I pay something like $15 to go from my home to my office in the morning. Something like $10 of that ride is going to the driver. If the ride is automated, either the price goes to $5, saving me $10 a ride which then is surplus to me, or the profit that Uber is making goes up significantly, which is surplus to them. Some of both is likely to happen. This surplus could be taxed, either at the company level or the individual level, so that the cost of the ride doesn’t go down nearly as much and the driver can continue to compete with the robot or the driver can collect some basic income, funded by the robot tax, while they find a new line of work.

At least that is the idea.

I would not characterize myself as a proponent of a robot tax or a basic income. But I find these ideas interesting and worth studying, debating, discussing, and testing at a small scale to understand their impacts. We should absolutely be doing that.

What Is Going To Happen In 2017

Happy New Year Everyone. Yesterday we focused on the past, today we are going to focus on the future, specifically this year we are now in. Here’s what I expect to happen this year:

  • Trump will hit the ground running, cutting corporate and personal taxes, and eliminating the preferential treatment of carried interest capital gains. The stock market has already factored in these tax cuts so it won’t be as big of a boon for investors as might be expected, but the seven and half year bull market run will be extended as a result of this tax cut stimulus before being halted by rising rates and/or some boneheaded move by President Trump which seems inevitable. We just don’t know the timing of it. The loss of capital gains treatment on carried interest won’t hurt professional investors too much because the lower personal tax rates will take the sting out of it. In addition, corporations will use the lower tax rates as an excuse to bring back massive amounts of capital that have been locked up overseas, producing a cash surplus that will result in an M&A boom. This will lead to an even more fuel to the fire that is causing “old line” corporations to acquire startups.
  • The IPO market, led by Snapchat, will be white hot. Look for entrepreneurs and the VCs that back them to have IPO fever in 2017. I expect we will see more tech IPOs in 2017 than we have since 2000.
  • The ad:tech market will go the way of search, social, and mobile as investors and entrepreneurs concede that Google and Facebook have won and everyone else has lost. It will be nearly impossible to raise money for an online advertising business in 2017. However, there will be new players, like Snapchat, and existing ones, like Twitter, that succeed by offering advertisers a fundamentally different offering than Facebook and Google do.
  • The SAAS sector will continue to consolidate, driven by a trifecta of legacy enterprise software companies (like Oracle), successful SAAS companies (like Workday), and private equity firms all going in search of additional lines of business and recurring subscription revenue streams.
  • AI will be the new mobile. Investors will ask management what their “AI strategy” is before investing and will be wary of companies that don’t have one.
  • Tech investors will start to adopt genomics as an additional “information technology” investment category, blurring the distinction between life science and tech investors that has existed in the VC sector for the past thirty years. This will lead to a funding frenzy and many investments will go badly. But there will be big winners to be had in this sector and it will be an important category for VCs for the foreseeable future.
  • Google, Facebook, and to a lesser extent Apple and Amazon will be seen as monopolists by government and individuals in the US (as they have been for years outside the US). Things like the fake news crisis will make clear to everyone how reliant we have become on these tech powerhouses and there will be a backlash. It will be Microsoft redux and the government will seek remedies which will be futile. But as in the Microsoft situation, technology, particularly decentralized applications built on open data platforms (ie blockchain technology), will come to the rescue and reduce our reliance on these monopolies. This scenario will take years to play out, but the seeds have been sown and we will start to see this scenario play out in 2017.
  • Cyberwarfare will be front and center in our lives in the same way that nuclear warfare was during the cold war. Crypto will be the equivalent of bomb shelters and we will all be learning about private keys, how to use them, and how to manage them. A company will make crypto mainstream via an easy to use interface and it will become the next big thing.

These are my big predictions for 2017. If my prior track record is any indication, I will be wrong about more of this than I am right. The beauty of the VC business is you don’t have to be right that often, as long as you are right about something big. Which leads to going out on a limb and taking risks. And I think that strategy will pay dividends in 2017. Here’s to a new year and new challenges to overcome.

What Did And Did Not Happen In 2016

As has become my practice, I will end the year (today) looking back and start the year (tomorrow) looking forward.

As a starting point for looking back on 2016, we can start with my What Is Going To Happen In 2016 post from Jan 1st 2016.

Easy to build content (apps) on a cheap widespread hardware platform (smartphones) beat out sophisticated and high resolution content on purpose built expensive hardware (content on VR headsets). We re-learned an old lesson: PC v. mainframe and Mac; Internet v. ISO; VHS v. Betamax; and Android v. iPhone.

And Fitbit proved that the main thing people want to do with a computer on their wrist is help them stay fit. And yet Fitbit ended the year with its stock near its all time low. Pebble sold itself in a distressed transaction to Fitbit. And Apple’s Watch has not gone mainstream two versions into its roadmap.

  • I thought one of the big four (Apple, Google, Facebook, Amazon) would falter in 2016. All produced positive stock performance in 2016. None appear to have faltered in a huge way in 2016. But Apple certainly seems wobbly. They can’t make laptops that anyone wants to use anymore. It’s no longer a certainty that everyone is going to get a new iPhone when the new one ships. The iPad is a declining product. The watch is a mainstream flop. And Microsoft is making better computers than Apple (and maybe operating systems too) these days. You can’t make that kind of critique of Google, Amazon, or Facebook, who all had great years in my book.
  • I predicted the FAA regulations would be a boon to the commercial drone industry. They have been.
  • I predicted publishing inside of Facebook was going to go badly for some high profile publishers in 2016. That does not appear to have been the case. But the ugly downside of Facebook as a publishing platform revealed itself in the form of a fake news crisis that may (or may not) have impacted the Presidential election.
  • Instead of spinning out HBO into a direct Netflix competitor, Time Warner sold itself to AT&T. This allows AT&T to join Comcast and Verizon in the “carriers becoming content companies” club. It seems that the executives who run these large carriers believe it is better to use their massive profits in the carrier business to move up the stack into content instead of continuing to invest in their communications infrastructure. It makes me want to invest in communications infrastructure honestly.
  • Bitcoin found no killer app in 2016, but did find itself the darling of the trader/speculator crowd, ending the year on a killer run and almost breaking the $1000 USD/BTC level. Maybe Bitcoin’s killer app is its value and/or store of value. That would make it the digital equivalent of gold and the likely reserve currency of the digital asset space. And I think that is what has happened with Bitcoin. And there is nothing wrong with that.
  • Slack had a good year in 2016, solidifying its position as the leading communications tool for enterprises (other than email of course). It did have some growing pains as there was a fair bit of executive turmoil. But I think Slack is here to stay and I think they can withstand the growing competition coming from Microsoft’s Teams product and others.
  • I was right that Donald Trump would get the Republican nomination and that the tech sector (with the exception of Peter Thiel and a few other liked minded people) would line up against him. It did not matter. He won the Presidency without the support of the tech sector, but by using its tools (Twitter and Facebook primarily) brilliantly.
  • I predicted “markdown mania” would hit the tech sector hard and employees would start getting cold feet on startups as they saw the value of their options going down. None of this really happened in a big way in 2016. There was some of that and employees are certainly more attuned to how they can get hurt in a down round or recap, but the tech sector has also used a lot of techniques, including repricing options, reloading option plans, and moving to RSUs, to mitigate this. The truth is that startups, venture capital, and tech growth companies had a pretty good year in 2016 all things considered.

So that’s the rundown on my 2016 predictions. I would give myself about a 50% hit rate. Which is not great but not horrible and about the same as I did last year.

Some other things that happened in 2016 that are important and worth talking about are:

  • The era of cyberwars are upon us. Maybe we have been fighting them silently for years. But we are not fighting them silently any more. We are fighting them out in the open. I suspect there is a lot that the public still doesn’t know about what is actually going on in this area. We know what Russia has done in the Presidential election and since then. But what has the US been doing to Russia? I would assume the same and maybe more. If your enemy has the keys to your castle, you had better have the keys to their castle. And as good as the Russians are at hacking into systems, the US has some great hackers too. I am very sure about that.  And so do the Chinese, the Israelis, the Indians, the British, the Germans, the French, the Japanese, etc, etc.  This feels a bit like the Nuclear era redux. Mutually assured destruction is a deterrent as long as both sides have the same tools.
  • The tech sector is no longer the belle of the ball. It has, on one hand become extremely powerful with monopolies, duopolies, or nearly so in search, social media, ecommerce, online advertising, and mobile operating systems. And it has, on the other hand, proven that it is susceptible to the very kinds of bad behavior that every other large industry is capable of. And we now have an incoming President who doesn’t share the love of the tech sector that our outgoing President showed. It brings to mind that scene in 48 Hours where Eddie Murphy throws the shot glass through the mirror and explains to the rednecks that there is a new sheriff in town. But this time, the tech sector are the rednecks.
  • Google and Facebook now control ~75% of the online advertising market and almost all of its growth in 2016:

  • Artificial Intelligence has inserted itself into our every day lives. Whether its a home speaker system that we can talk to, or a social network that already knows what we are about to go out and purchase, or a car that can park itself and change lanes on the highway automatically, we are seeing AI take over tasks that we used to have to do ourselves. We are in the age of AI. It is not something that is coming. It is here. It may have arrived in 2014, or 2015, but if you ask me, I would put 2016 as the year it had its debut in mainstream life. It is exciting and it is scary. It begs all sorts of questions about where we are all going in the next thirty to fifty years. If you are in your twenties, AI will define your lifetime.

So that’s my rundown on 2016. I wish everyone a happy and healthy new year and we will talk about the future, not the past, tomorrow.

If you are in need of a New Year’s Resolution, I suggest moving to super secure passwords and some sort of tool to manage them for you, using two factor authentication whenever and wherever possible, encrypt as much of your online activities as you reasonably can, and not saying or doing anything online that you would not do in public, because that is where you are doing it.

Happy New Year!

Data Wins

Whenever people ask me which company I think will win the self driving car race, I say Tesla.

And the reason is that they have more data.

And when it comes to training machines to do what humans do, more data is better than more software engineers.

Bloomberg has a good post on that today.

Headlines

One of the issues in all of the concerns about “fake news” is the way headlines are used on the Internet. Newspapers and magazines certainly took the construction of headlines into account to drive readers into the stories. But on the Internet, headlines have become that and more. They are the links themselves that fly around the Internet and “convert” someone into coming to your site and reading a story. They are “clickbait.” If we want to address the veracity and authenticity of content on the Internet, we might want to start with headlines.

I’ve had my issues with headlines for years. Many years ago, I allowed a number of publications to repost content I write here at AVC on their online publications. The publication that does that most frequently with my content is Business Insider. You can see the hundreds of posts that BI has republished on my author page at Business Insider. When they started doing this maybe seven or eight years ago, I would notice that they would leave my post intact, verbatim, but rewrite the headline. It would drive me crazy because I view the headline as an integral part of my post. I think about the words I use to title my posts. So I would send them angry emails and most of the time they would change it back. But it was a lesson in the difference between a headline that I liked and a headline that would drive clicks.

I also have seen hundreds of stories written about me, USV, and our portfolio companies that have sensational and often inaccurate headlines followed by stories that are essentially correct and well reported. It drives me nuts but I don’t often do much about it.

It makes me think that someone, or some company, or some open source community ought to build software that parses headlines and the stories that follow and rate them for how well the headline represents the article. That “headline veracity ranking” could then be offered to anyone who presents headlines to readers. That would be social media like Facebook, Twitter, Reddit, etc. That would be email applications and browsers. That would be search engines. Etc, etc, etc.

It would be nice to see some competition in this sector so that one company doesn’t become the arbiter of what is an accurate headline and what is not. That doesn’t sound like a good outcome. But if this is done via open source, or is community powered in some way, this could be a very helpful tool in getting publishers to behave and represent their stories accurately.

And that would be a wonderful thing for the Internet.

AI: Why Now?

UK-based VC David Kelnar wrote an excellent primer on Artificial Intelligence that is a relatively quick read and helps explain the technology and its advancement over the past sixty years since the term was coined in the mid 1950s.

I like this chart which explains the relationship between AI, machine learning, and deep learning.

But my favorite part of David’s post is his explanation of why AI has taken off in the past five years, as this chart shows:

Like most non-linear curves, it is not one thing, but a number of things happening simultaneously, that is causing this explosion of interest. David cites four things:

  1. Better algorithms. Research is constantly coming up with better ways to train models and machines.
  2. Better GPUs. The same chips that make graphics come alive on your screen are used to train models, and these chips are improving rapidly.
  3. More data. The Internet and humanity’s use of it has produced a massive data set to train machines with.
  4. Cloud services. Companies, such as our portfolio company Clarifai, are now offering cloud based services to developers which allow them to access artificial intelligence “as a service” instead of having to “roll your own”.

I feel like we are well into the “AI wave” of technology right now (following in order web, social, and mobile) and this is a wave that seemingly benefits the largest tech companies like Google, Facebook, Amazon, Microsoft, IBM, Uber, Tesla which have large datasets and large userbases to deploy this technology with.

But startups can and will play a role in this wave, in niches where the big companies won’t play, in the enterprise, and in building tech that will help deliver AI as a service. David included this chart that shows the massive increase in startup funding for AI in the last four years:

I would like to thank David for writing such a clear and easy to understand primer on AI. I found it helpful and I am sure many of you will too.